SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

addressing the issue #34611 to make FlaxDinov2 compatible with any batch size (#35138)

Browse Source 
fixed the batch_size error, all tests are passing

Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com>
This commit is contained in:
MAHIR DAIYAN
2025-02-25 18:44:44 +08:00
committed by GitHub
parent 3a02fe56c2
commit d80d52b007
2 changed files with 23 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/transformers/models/dinov2/modeling_flax_dinov2.py
View File

@@ -185,9 +185,11 @@ class FlaxDinov2Embeddings(nn.Module):
antialias=False, antialias=False,
) )
patch_pos_embed = patch_pos_embed.astype(target_dtype) patch_pos_embed = patch_pos_embed.astype(target_dtype)
patch_pos_embed = jnp.transpose(patch_pos_embed, (0, 2, 3, 1)).reshape((hidden_states.shape[0], -1, dim)) patch_pos_embed = jnp.transpose(patch_pos_embed, (0, 2, 3, 1)).reshape((position_embeddings.shape[0], -1, dim))
patch_pos_embed_expanded = jnp.tile(patch_pos_embed, (hidden_states.shape[0], 1, 1))
class_pos_embed_expanded = jnp.tile(class_pos_embed, (hidden_states.shape[0], 1, 1))
return jnp.concatenate((class_pos_embed[jnp.newaxis, :], patch_pos_embed), axis=1) return jnp.concatenate((class_pos_embed_expanded, patch_pos_embed_expanded), axis=1)
def __call__(self, pixel_values, deterministic=True): def __call__(self, pixel_values, deterministic=True):
batch_size = pixel_values.shape[0] batch_size = pixel_values.shape[0]
@@ -778,7 +780,7 @@ FLAX_VISION_CLASSIFICATION_DOCSTRING = """
>>> image = Image.open(requests.get(url, stream=True).raw) >>> image = Image.open(requests.get(url, stream=True).raw)
>>> image_processor = AutoImageProcessor.from_pretrained("facebook/dinov2-base-imagenet1k-1-layer") >>> image_processor = AutoImageProcessor.from_pretrained("facebook/dinov2-base-imagenet1k-1-layer")
>>> model = FlaxDinov2ForImageClassification.from_pretrained("facebook/dinov2-base-imagenet1k-1-layer") >>> model = FlaxDinov2ForImageClassification.from_pretrained("facebook/dinov2-base-imagenet1k-1-layer", from_pt=True)
>>> inputs = image_processor(images=image, return_tensors="np") >>> inputs = image_processor(images=image, return_tensors="np")
>>> outputs = model(**inputs) >>> outputs = model(**inputs)

28
tests/models/dinov2/test_modeling_flax_dinov2.py
View File

@@ -202,7 +202,7 @@ class FlaxDionv2ModelTest(FlaxModelTesterMixin, unittest.TestCase):
# We will verify our results on an image of cute cats # We will verify our results on an image of cute cats
def prepare_img(): def prepare_img():
image = Image.open("./tests/fixtures/tests_samples/COCO/000000039769.png") image = Image.open("./tests/fixtures/tests_samples/COCO/000000039769.png")
return image return [image, image]
@require_vision @require_vision
@@ -224,18 +224,25 @@ class FlaxDinov2ModelIntegrationTest(unittest.TestCase):
outputs = model(pixel_values=pixel_values) outputs = model(pixel_values=pixel_values)
# verify the logits # verify the logits
expected_shape = (1, 257, 768) expected_shape = (2, 257, 768)
self.assertEqual(outputs.last_hidden_state.shape, expected_shape) self.assertEqual(outputs.last_hidden_state.shape, expected_shape)
expected_slice = np.array( expected_slice = np.array(
[ [
[-2.1629121, -0.46566057, 1.0925977], [
[-3.5971704, -1.0283585, -1.1780515], [-2.1629121, -0.46566057, 1.0925977],
[-2.900407, 1.1334689, -0.74357724], [-3.5971704, -1.0283585, -1.1780515],
[-2.900407, 1.1334689, -0.74357724],
],
[
[-2.1629121, -0.46566057, 1.0925977],
[-3.5971704, -1.0283585, -1.1780515],
[-2.900407, 1.1334689, -0.74357724],
],
] ]
) )
self.assertTrue(np.allclose(outputs.last_hidden_state[0, :3, :3], expected_slice, atol=1e-4)) self.assertTrue(np.allclose(outputs.last_hidden_state[:2, :3, :3], expected_slice, atol=1e-4))
@slow @slow
def test_inference_image_classification_head_imagenet_1k(self): def test_inference_image_classification_head_imagenet_1k(self):
@@ -252,12 +259,13 @@ class FlaxDinov2ModelIntegrationTest(unittest.TestCase):
logits = outputs.logits logits = outputs.logits
# verify the logits # verify the logits
expected_shape = (1, 1000) expected_shape = (2, 1000)
self.assertEqual(logits.shape, expected_shape) self.assertEqual(logits.shape, expected_shape)
expected_slice = np.array([-2.1776447, 0.36716992, 0.13870952]) expected_slice = np.array([[-2.1776447, 0.36716992, 0.13870952], [-2.1776447, 0.36716992, 0.13870952]])
self.assertTrue(np.allclose(logits[0, :3], expected_slice, atol=1e-4)) self.assertTrue(np.allclose(logits[:2, :3], expected_slice, atol=1e-3))
expected_class_idx = 281 expected_class_idx = 281
self.assertEqual(logits.argmax(-1).item(), expected_class_idx) self.assertEqual(logits[0].argmax(-1).item(), expected_class_idx)
self.assertEqual(logits[1].argmax(-1).item(), expected_class_idx)